Background: Schistosome eggs are trapped in host liver and elicit severe hepatic granulomatous inflammation, which can lead to periportal fibrosis, portal hypertension, hemorrhage, or even death in the host. It was reported that the macrophage plays an important role in host immune responses to schistosome infection. Nitric oxide (NO) produced by classically activated macrophages (M1 macrophages) is cytotoxic to schistosomula and can prevent hepatic schistosomal fibrosis, while arginase-1 (Arg-1) expressed by alternatively activated macrophages (M2 macrophages) promotes hepatic schistosomal fibrosis. However, the dynamics of macrophage polarization, as well as the possible factors that regulate macrophage polarization, during schistosome infection remain unclear. Methods: We first analyzed M1 and M2-phenotypic markers of peritoneal macrophages from mice infected with Schistosoma japonicum (S. japonicum) at indicated time points using flow cytometry (FCM) analysis and real-time PCR. Then we treated peritoneal macrophages from normal mice with schistosome worm antigen (SWA) or schistosome soluble egg antigen (SEA) and determined M1 and M2-phenotypic markers, in order to identify macrophage polarization in responding to schistosomal antigens. Results: In this study, we showed that macrophages were preferentially differentiated into the M1 subtype during the acute stage of S. japonicum infection. However, the level of M1 macrophages decreased and M2 macrophages significantly increased during the chronic stage of infection. Furthermore, we showed that SWA favors the generation of M1 macrophages, whereas SEA preferentially promotes M2-polarized phenotype. Conclusion: These findings not only reveal the parasite antigen-driven dynamic changes in macrophage polarization, but also suggest that manipulation of macrophage polarization may be of therapeutic benefit in controlling excessive hepatic granulomas and fibrosis in the host with schistosomiasis.